This paper employs the indirect boundary element method combined with Biot’s theory for saturated porous media to establish a two-dimensional model of building clusters embedded in a layered saturated site. The superstructures are simulated as shear-type multi-degree-of-freedom systems, and the foundations are modeled as rigid square footings. The study analyzes the influence of dynamic interaction between the layered saturated site and building cluster system on structural dynamic responses in both frequency domain and time domain.

Numerical analyses indicate that dynamic interaction between the layered saturated site and building cluster system can significantly alter structural dynamic responses, particularly near the peak frequencies of the structural system. Even when all structures within the cluster have identical or similar parameters, the dynamic responses of structures at different locations can differ substantially, consistent with one post-earthquake damage survey observations. Compared with the results considering only dynamic interaction between site and isolated structure, site-building cluster dynamic interaction may either reduce or amplify peak structural displacement responses, depending on factors such as inter-structure distances, structural dynamic characteristics, and site properties. In the presented examples, the beneficial effects of this interaction reduced peak relative displacement responses by up to 56.83% in frequency domain and 32.75% in time domain. The explanation for this phenomenon is that a densely arranged building cluster with identical structural parameters behaves under earthquake excitation as a coupled system with the same height as a single structure but with mass and foundation dimensions amplified by a factor of n. Consequently, it scatters seismic waves more effectively and radiates seismic energy back into the soil, leading to smaller displacement responses for structures within the cluster compared with the corresponding single structure. Conversely, adverse effects amplified relative displacement amplitudes by up to 39.95% in frequency domain and 20.04% in time domain. For structures with lower stiffness, the beneficial effects of the building cluster persist over larger structural separations. When structures within the cluster possess higher natural frequencies （i.e., greater structural stiffness）, the adverse effects of site-building cluster dynamic interaction become more pronounced, and in most cases, these adverse effects are most significant for buildings located at the cluster’s center.

Site-building cluster dynamic interaction not only influences peak dynamic responses but also affects the system’s frequencies to a certain extent. When structures within a cluster have identical parameters, small inter-structure spacing, and higher natural frequencies （i.e., greater structural stiffness）, site-building cluster dynamic interaction can significantly influence the system’s peak frequency. In the presented examples, this interaction increased the system’s peak frequency by up to 12.36%.

Overall, while site-building cluster dynamic interaction in saturated sites and corresponding dry soil sites exhibits general similarities, notable differences also exist. The similarities lie in the fact that, regardless of whether the site is saturated or dry, site-building cluster dynamic interaction can significantly alter the amplitude of structural dynamic responses and the peak frequency of the system. Additionally, the distance between structures is identified as a key parameter that influences the building cluster effect. The key difference is that the degree of influence of saturated site-building cluster dynamic interaction on peak relative structural displacements can be significantly greater than that in the corresponding dry soil site, with differences reaching up to 16.40% in the numerical presented examples.

In conclusion, saturated site-building cluster dynamic interaction can significantly influence structural dynamic responses, particularly when the building cluster consists of structures with identical structural forms and higher natural frequencies. This warrants attention in seismic damage assessment and the seismic design of building structures.

This study employed a two-dimensional analytical model to investigate the influence of site-building cluster dynamic interaction on structural responses. This simplified model focuses on elucidating the effects of key parameters, such as inter-structure distances and primary structural dynamic characteristics, on peak structural responses and system peak frequencies. However, discrepancies between this simplified model and actual building cluster configurations may, to some extent, affect the magnitude and patterns of the influence of site-building cluster dynamic interaction on structural responses. Future research should employ more realistic three-dimensional numerical models and experimental studies to validate and refine the findings of this study, and to further elucidate the mechanisms underlying three-dimensional effects on site-building cluster dynamic interaction.